PM_2.5 Source Variation And Spatial-temporal Distribution Based On LUR Model In Xi'an

With the rapid development of Chinese economy and the accompanying quick growth of energy consumption,air pollution incidents are increasing.Near the ground,high concentrations of air pollutants can directly threaten human health,but also seriously restrict the smooth development of social economy.Therefore,the Chinese government attaches great importance to the analysis and prevention of influencing factors of air pollution.Most of the research on air pollution in China has been concentrated in developed areas,but less on western cities.In addition,with the implementation of the policy of"The Belt and Road",more and more investment has been brought to the west,and the rapid economic growth will inevitably have a profound impact on the atmospheric environment.Xi'an is located in the North China-Guanzhong Plain,one of the major air pollution areas in China.Due to the double restrictions of natural and human factors,its air quality has not been able to meet the air quality standards of China.At the same time,China has entered a new stage of development since the 21st century,and the public has put forward higher requirements for environmental quality,especially for atmospheric environmental quality.Therefore,we took Xi'an as the target city,using the hourly concentration values of five air pollutants(including particulate matter:PM_2.5,PM₁₀ and gaseous pollutants:SO₂,CO,NO₂)at 10monitoring stations from January 1,2016 to December 31,2016 in the urban area of Xi'an.The temporal variation characteristics of five air pollutants are analyzed,which are combined with the clustering results of air mass backward trajectories and satellite images in spring and summer.The contribution of PM_2.5 combustion/non-combustion source and its variation characteristics were obtained.On this basis,46 predictive variables are selected to fit the LUR model.Considering that Xi'an is an important tourist city as well as a science and education center,and its urban functional zones are diverse.This paper divides the urban area of Xi'an into five functional zones?factory,residence,parks,tourist spots and universities?,calculates the spatial distribution of PM_2.5 concentration in different functional zones,and then discusses its healthy impact on residents of different age levels?infants,children and adults?.The following main conclusions are drawn:?1?During the research period,PM_2.5,as the main pollutant in Xi'an,exceeded the national air quality secondary standard for 114 days,mainly in winter and spring.The annual PM_2.5 concentration showed a single peak"U"pattern,with the highest concentration in winter?120.8?g/m3?,the second in autumn?80.5?g/m3?,spring?61.8?g/m3?,and the lowest in summer?32.4?g/m3?.The daily concentration of PM_2.5 in Xi'an is of double peaks and double valleys.The peak value is mainly caused by the morning and evening travel peak.With the end of the morning peak,the concentration of particulate matter gradually decreased from 12:00 to a low point.The arrival of the late peak,together with the possible simultaneous occurrence of night breeze or calm wind and radiation inversion at night,causing PM_2.5 concentration to rise again and reach its peak at 23:00-24:00.?2?This paper divides the whole year into cold season?November 1-March 31?and warm season?April 1-October 31?.It was found that the contribution rates of combustion sources in cold and warm seasons in Xi'an were 22.7%and 36.8%,respectively.The high value of PM_2.5 non-combustion source appeared from March to June,and decreased slightly in July until December.The day when the average PM_2.5concentration is more than 75?g/m3is defined as pollution day.The correlation between PM_2.5 concentration and SO₂,CO and NO₂ concentration in Xi'an is good on both pollution day and non-pollution day.Besides SO₂,the coefficient of pollution period in Xi'an is larger than that in Non-pollution period,which indicates that the pollution of PM_2.5 in Xi'an is mainly caused by human sources.?3?LUR model is considered to be one of the suitable methods for predicting the spatial variation of air pollution concentration.In this paper,five predictive variables are introduced into the model prediction.A total of 98.1%of the model changes are explained,and the deviation and root mean square error are small.Compared with the traditional interpolation method,LUR model not only has higher simulation accuracy,but also has better spatial heterogeneity expression because it can explain the spatial variability and non-stationarity of the predicted variables.In order to study the impact of PM_2.5 on human health,we divided the urban area of Xi'an into 3494 points according to the 500 m*500 m grid,simulated the spatial distribution of PM_2.5 concentration in Xi'an?divided five functional areas?.Children?8-10 years old?had the highest health risk?4.78?;adults?2.58?had the lowest health risk,and infants?2.72?were between the two.